Rate forecasts are projections, not guarantees; all figures reflect data published as of September 2026 and are for general information, not personalized financial advice.
TL;DR — Quick Verdict
- Freddie Mac put the 30-year fixed-rate mortgage at 6.76% on September 10, 2026 — up from 6.71% the week before and 6.66% two weeks before that, the highest weekly average recorded so far in 2026.
- Rates are now further above where they stood a year ago, when the 30-year averaged 6.35%. The 2026 story reversed in late February, when the outbreak of war with Iran pushed energy costs and inflation expectations higher, and that reversal has only deepened as the conflict has ground on into September.
- Four major forecasters’ full-year 2026 estimates still range from roughly 6.18% to 6.5% — but that apparent stability masks a real shift: Fannie Mae raised its own outlook by two-tenths of a point in a single month, converging with the Mortgage Bankers Association’s already-cautious view. Every published annual average still sits below the 6.76% the market is charging today.
- Waiting a year for a 0.5-point rate improvement on a $343,280 loan saves roughly $113 per month — but NAR reports the median existing-home price rose 1.6% year over year to $429,100, which erases nearly a third of that gain if prices repeat.
- The 30-year fixed-rate mortgage now sits about 196 basis points above the 10-year Treasury yield of 4.80%, wider than the historical norm near 170 basis points.
Four institutions published 30-year mortgage rate forecasts for 2026 within the last few months. Their full-year projections still range from roughly 6.18% to 6.5% — a spread of about 32 basis points. But that number hides more movement than it shows: Fannie Mae’s own forecast jumped from 6.3% to 6.5% in a single month, now matching the Mortgage Bankers Association’s outlook almost exactly, while the low end of the range comes from a National Association of Home Builders projection that hasn’t been revised since June and increasingly looks disconnected from where the market has gone since.
Freddie Mac’s Primary Mortgage Market Survey recorded 6.76% on September 10, 2026, up from 6.71% the week before and 6.66% the week before that — three straight weekly increases that pushed the rate to its highest weekly average of the year. Rate-shopping borrowers at Rocket Mortgage or a local credit union will see quotes swing more than the current five-basis-point weekly move in a single afternoon.
This analysis breaks down what each major forecaster actually projects, models the dollar cost of waiting for lower rates against rising home prices, and identifies which borrowers benefit from patience and which lose money by exercising it. Every figure traces to Freddie Mac, Fannie Mae, the Mortgage Bankers Association, the Federal Reserve, the Federal Housing Finance Agency, or the National Association of Realtors.
What Every Major Forecaster Projects Through 2028
Consensus is eroding faster than the headline numbers suggest. Fannie Mae’s Economic and Strategic Research Group, the Mortgage Bankers Association, the National Association of Home Builders, and Wells Fargo’s economics team still land within roughly a third of a percentage point of each other for full-year 2026 — but that’s largely because Fannie Mae just moved sharply toward the others rather than because the group is converging on a stable outlook.
Sources: Fannie Mae Economic and Strategic Research Group August 2026 Housing Forecast, Mortgage Bankers Association July 2026 Mortgage Finance Forecast, National Association of Home Builders, Wells Fargo U.S. Economic Outlook (verify at fanniemae.com, mba.org, nahb.org). *NAHB has not published a revision since June and predates two additional months of rate increases; treat as stale. †Wells Fargo’s August update gave Q3/Q4 2026 and full-2027 quarterly figures rather than a restated annual average; a complete 2026 annual figure could not be confirmed for this update.
Read those numbers against the 6.76% currently on offer and the picture keeps shifting in one direction. Fannie Mae’s Q4 2026 forecast — 6.8% — is now actually higher than today’s spot rate, a reversal from earlier this year when every forecast trailed the market on the way up. The forecast community isn’t predicting rates fall from here so much as one major forecaster now expects them to climb a bit further before this cycle turns. Fannie Mae’s own note ties its August outlook to market levels from around mid-month, before the rate’s continued push to new highs into September, so even this sharply revised figure may still be lagging.
NAHB remains the outlier on the optimistic side, and its own economist has previously qualified the projection — the group does not expect the 30-year fixed-rate mortgage to sit consistently below 6% until late 2027. That caveat matters more than ever now: the figure predates the market’s climb from the mid-6.60s to 6.76% and hasn’t been tested against it.
Forecast revisions also reveal direction, and this one was steep. Fannie Mae’s June and July outlooks both had rates holding near 6.3% through 2026 and 2027. The August revision jumped the group’s Q3 2026 estimate to 6.7%, Q4 to 6.8%, and pushed 2027 up to an average of 6.7% — one of the sharpest one-month upward revisions the forecast has made this cycle. Anyone tracking current 30-year fixed rate trends week to week should treat these revisions as more informative than the point estimates themselves.
Why the Fed Is the Wrong Thing to Watch
Borrowers ask about Federal Reserve meetings. They should be asking about oil and Treasury auctions.
Mortgage rates are priced off the 10-year Treasury yield, which stood at 4.80% on September 10, 2026, plus a risk premium demanded by mortgage-backed securities investors. With the 30-year fixed-rate mortgage at 6.76%, that premium — the spread — runs roughly 196 basis points. Historically the spread has averaged closer to 170 basis points, meaning today’s borrower pays about 0.26 percentage points more than Treasury math alone would dictate.
That excess is worth real money. On a $343,280 loan — 80% of the $429,100 national median existing-home price reported by NAR for August 2026 — closing that gap would save roughly $59 per month, or about $21,250 across a full 30-year term. Spread compression, not Fed policy, is the mechanism most likely to deliver meaningful relief. The relationship between Fed policy and mortgage rates is indirect enough that the two have moved in opposite directions during recent cutting cycles.
The July 29 FOMC meeting demonstrated the point. The Committee voted 9–3 to hold the federal funds target range at 3.50%–3.75%, where it has stood since December 2025. What moved markets was not the hold but the dissent: Cleveland’s Beth Hammack, Minneapolis’ Neel Kashkari and Dallas’ Lorie Logan each preferred a quarter-point increase. Long-term yields rose in the weeks that followed, carrying mortgage rates past their earlier 2026 peak and toward the fresh highs recorded in early September.
Chair Kevin Warsh has also changed how the Fed communicates. Post-meeting statements under his leadership are substantially shorter, and he has stripped forward guidance out of them entirely, arguing that markets should read the data rather than the Committee’s projections. For borrowers the practical effect is more volatility between meetings, not less — with no guidance to anchor expectations, each inflation and jobs print moves yields further than it used to.
The underlying driver is still energy, and it has intensified. The war with Iran that began at the end of February has continued into September, with oil trading above $100 a barrel in recent sessions as the conflict shows no clear sign of resolution. The August CPI report, released September 11, showed headline inflation accelerating to 0.4% for the month — the fastest pace in three months — driven largely by a 3.9% monthly jump in gasoline prices, up 27.4% year over year. The annual headline rate held at 3.4%. Core inflation actually ticked up on a monthly basis, to 0.3% from July’s 0.2%, though the annual core rate eased to 2.4%, its lowest level since March 2021, as shelter costs continued to cool. Following the report, fed funds futures and prediction markets pushed the odds of a quarter-point hike at this Wednesday’s meeting to roughly 90%, up from about 70% beforehand — a dramatic shift from the mere hawkish dissent the Committee logged in July. As of this writing the meeting hasn’t happened yet; if the hike materializes, it would be the first increase of this cycle. The Committee also releases updated economic projections at that meeting.
Waiting for Lower Rates vs. Buying Now: Which Costs Less?
Assume you are shopping at the national median price of $429,100 with 20% down, financing $343,280. Today’s 6.76% produces a principal-and-interest payment near $2,229. Now model the wait.
Modeled by Real Cost Report using the standard amortization formula. Rate inputs from Freddie Mac PMMS and the Fannie Mae ESR Group August 2026 Housing Forecast; price appreciation from the National Association of Realtors August 2026 Existing-Home Sales report (verify at freddiemac.com, fanniemae.com, nar.realtor).
Half a point of rate improvement saves $113 monthly in isolation. Layer in a repeat of the 1.6% annual price gain NAR recorded and the saving shrinks to $79 — a nearly one-third reduction. Should Fannie Mae’s own newly bearish Q4 2026 projection prove accurate instead, the waiting buyer pays $45 more per month than the buyer who acted today, plus twelve months of rent with no equity accrual. Notably, the MBA’s flatter 6.5% forecast — once the “pessimistic” scenario in this comparison — would now actually beat today’s rate by about $24 a month with the same price appreciation, since the market has moved further above MBA’s number than home prices have appreciated. Which scenario is realistic now depends more on which forecaster you trust than it did a month ago.
Notice what the optimistic scenarios require. A move to 6.26% within twelve months would undercut every published 2026 and 2027 forecast in the table above, including Fannie Mae’s own newly-raised 6.7% average for 2027. It is possible — spread compression alone could deliver a meaningful chunk of it — but it is not the base case any institution is currently publishing.
Verdict
Waiting still lowers the monthly payment in most modeled scenarios, but the range has widened: anywhere from $79 in savings to $45 in extra cost, depending on which forecaster’s outlook plays out — and for the first time this cycle, one of the major forecasters (Fannie Mae) now expects things to get modestly worse before they get better. Against that sits twelve months of rent with no equity accrual and a price trend that has now produced 38 consecutive months of year-over-year increases. Buyers whose finances support the payment today should transact and treat a future refinance as a free option; buyers stretching to qualify should wait for income growth, not rate relief.
What Most People Get Wrong About Rate Forecasts
Four errors show up repeatedly in how borrowers translate forecasts into action.
Mistake 1: Treating an annual average as a rate you can lock
Fannie Mae’s 6.5% annual figure for 2026 — revised sharply upward just last month — is still a blend of quarterly projections, and its rate path was built off market levels from mid-August, before rates climbed further. No lender offers “the annual average.” The consequence is a borrower who declines a 6.7% lock waiting for 6.5% and ends up at 6.76% or higher. Correct action: compare live quotes to the forecast for the specific quarter you plan to close in, and understand how rate lock timing and extension costs work before you gamble on a few basis points.
Mistake 2: Assuming the advertised rate applies to you
Freddie Mac’s survey covers conventional, conforming, fully amortizing purchase loans for borrowers putting 20% down with excellent credit. Miss any of those conditions and your quote differs. The consequence is budgeting at 6.76% and receiving something meaningfully higher. Correct action: model your actual profile, factoring in credit score pricing tiers and your down payment.
Mistake 3: Ignoring loan size thresholds
FHFA set the 2026 baseline conforming loan limit at $832,750, up $26,250 from 2025, with a high-cost ceiling of $1,249,125. Financing one dollar above your county limit moves you into different pricing entirely. Correct action: check your county before making an offer and review jumbo versus conforming rate differences.
Mistake 4: Optimizing the rate while ignoring fees
A 6.5% quote with $8,000 in origination charges costs more than a 6.76% quote with $2,000. Consequence: borrowers chase headline rates into worse total-cost outcomes. Correct action: evaluate lenders on APR and total cash outlay, using APR-based lender comparison rather than rate alone, and price discount points and buydown math against your expected tenure in the home.
Who Should Act Now and Who Should Wait
Forecast uncertainty argues for decisions driven by personal circumstance rather than market timing. The following conditions separate the two groups cleanly.
Act now if: your housing payment including taxes and insurance stays under 30% of gross income at 6.76%; you hold six months of reserves after closing; you expect to stay seven years or more; or you are financing above the conforming limit, where jumbo pricing has recently run competitive with conforming and that relationship may not persist.
Wait if: your credit score sits within 20 points of a higher pricing tier — improving it delivers more certain savings than any forecast; you have less than 10% down and would carry mortgage insurance for years; your employment is under twelve months old at a new employer; or you are within eighteen months of a documented income increase.
Shorter-horizon buyers face a different calculation entirely. If your realistic ownership window is five to seven years, an adjustable-rate product may beat a fixed rate outright — run the ARM versus fixed break-even analysis before assuming otherwise, and compare against 15-year and 30-year total interest outcomes. The 15-year averaged 6.09% in the same September 10 survey, a 67-basis-point discount to the 30-year. Veterans and eligible service members should separately price VA loan rates against conventional, since the no-down-payment structure changes the waiting calculus substantially.
One universal action applies regardless of category: obtain quotes from at least three lenders on the same day. Rate dispersion across lenders routinely exceeds the entire 32-basis-point spread separating the 2026 forecasts, which means the controllable factors behind your rate matter more than what any economist projects.
Frequently Asked Questions
Will mortgage rates drop below 6% in 2026?
No major forecaster projects a sustained sub-6% average for 2026, and the market has moved further away from that outcome since these forecasts were published. Fannie Mae itself now projects a 6.5% full-year average and 6.8% for the fourth quarter — higher than several other forecasters’ entire-year estimates. The most optimistic institutional figure still on record comes from NAHB at 6.18%, but that projection hasn’t been updated since June and predates the market’s climb to 6.76%. Rates did briefly fall below 6% in late February 2026, before the outbreak of war with Iran reversed the trend — a reversal that has continued to deepen into September.
How accurate have these forecasts been historically?
Directionally reasonable, precisely unreliable. Fannie Mae’s September 2025 outlook projected rates ending 2026 at 5.9%; the actual September 10, 2026 reading from Freddie Mac was 6.76%. The gap of roughly 0.86 percentage points on a $343,280 loan equals about $193 monthly. Treat forecasts as scenario inputs, not commitments.
Should I buy points if rates are forecast to fall?
Generally no, if you believe the forecasts — and the case has only gotten stronger. Points typically require five to seven years to break even, and refinancing before break-even forfeits the entire upfront cost. Fannie Mae’s own forecast now shows just a six-basis-point improvement by late 2027, from 6.76% today to roughly 6.7%. Paying for a permanent buydown you may abandon within two years rarely produces positive expected value against an improvement that small. Each quarter point of permanent buydown is worth about $57 per month at the median loan size.
Does the 10-year Treasury predict mortgage rates reliably?
It sets the floor, not the price. With the 10-year at 4.80% on September 10, 2026, and the 30-year fixed-rate mortgage at 6.76%, the spread runs roughly 196 basis points against a historical norm nearer 170. Watching Treasury moves tells you the direction of pressure; spread behavior determines how much of that move reaches borrowers.
Are rates higher or lower than a year ago?
Higher, and by a wider margin than earlier this year. The 30-year averaged 6.35% in early September 2025 against 6.76% today, and the 15-year averaged 5.50% against 6.09%. Affordability has still improved on NAR’s measure, because income growth has outrun the rate increase — the association’s Housing Affordability Index read 104.7 in August versus 101.2 a year earlier, an improvement recorded across all four regions. Details are available in Freddie Mac’s Primary Mortgage Market Survey.
How We Researched This Article
Every rate figure in this analysis was drawn from primary institutional publications retrieved in September 2026. Current market rates come from the Freddie Mac Primary Mortgage Market Survey, which compiles mortgage rates from loan applications submitted through Loan Product Advisor and publishes weekly on Thursdays. The survey covers conventional, conforming, fully amortizing home purchase loans for borrowers with 20% down and excellent credit — a narrower population than the market as a whole, which is why individual quotes commonly exceed the published average.
Forecast figures were taken from four sources: the Fannie Mae Economic and Strategic Research Group August 2026 Housing Forecast, the Mortgage Bankers Association Mortgage Finance Forecast as of July 2026, the National Association of Home Builders June 2026 market update, and the Wells Fargo U.S. Economic Outlook from August 2026. Where a forecaster publishes quarterly projections, we report the annual average the institution itself states rather than computing our own; for Wells Fargo, only Q3/Q4 2026 and 2027 quarterly figures could be confirmed for this update, so those are presented directly rather than as a restated annual average. NAHB’s figures have not been revised since June 2026 and predate the market’s subsequent climb; we retain them because no updated NAHB forecast could be sourced, but readers should treat them as the most dated numbers in the table.
Home price and transaction data come from the National Association of Realtors Existing-Home Sales report released September 10, 2026, covering August. Loan limit values come from the Federal Housing Finance Agency conforming loan limit announcement of November 25, 2025. Treasury yields were verified against Federal Reserve Economic Data series DGS10, sourced from Treasury market data for September 10, 2026. Monetary policy details come from the FOMC statement of July 29, 2026 — the most recent decision on record — and inflation figures from the Bureau of Labor Statistics Consumer Price Index release of September 11, 2026.
All payment figures are modeled, not measured. We applied the standard fixed-rate amortization formula to principal and interest only — property taxes, homeowners insurance, mortgage insurance, and HOA dues are excluded and will materially increase actual monthly outlay. The 20% down payment assumption aligns our scenarios with the Freddie Mac survey population; borrowers with smaller down payments will see different rates and additional costs. Spread comparisons against the historical 170-basis-point norm use a long-run approximation rather than a fixed measured series, and should be read as directional context.
Limitations worth naming: forecasts are revised monthly and several cited here will change before you read this. The FOMC meets again this Wednesday, September 16, 2026 — after this article was researched but likely before most readers see it — and markets are pricing a roughly 90% probability of a quarter-point hike, which would be the first increase of this cycle and could move mortgage rates meaningfully in either direction depending on the accompanying projections and statement language. NAHB’s figures in the forecast table predate two additional months of rate increases and should be treated with particular caution. Our price-appreciation scenario assumes the August 2026 year-over-year rate repeats, which is a modeling convenience rather than a prediction. Note also that the August median of $429,100 is below July’s $434,100 — a seasonal pattern, not a decline in the trend, which remains positive year over year for a 38th consecutive month. Research was last conducted September 2026. All figures were verified against named primary sources before publication.